The Man Who Lied to his Laptop by Clifford Ness with Corina Yen
According to the late Clifford Nass, “people treat computers and other interactive technologies like actual people.” Believing we humans operate on simple principles, Nass worked “at the intersection between social science and technology.” This Stanford professor and his colleagues carried out plenty of research to support these claims.
He has also worked with private industry to improve software designs and make them more acceptable to users. One challenge involved re-designing Microsoft’s obnoxious Clippy, one of the most hated interfaces in computer history.
By now, we’ve all heard multiple voices animating our GPS systems. We know these are not real people — just random voices that express what the computer program decides is the best route to take. But research shows that such voices are treated by many customers as if they were human.
For its Five Series cars, BMW was forced to recall a navigation system that was “well ahead of other companies in terms of accuracy and functionality.” Why? Because “male German drivers disliked taking directions from a woman, “flooded the service desk with complaints,” and refused to be mollified by the explanation that it was not a real woman. They knew this; it didn’t matter.
Experiments involving groups of people receiving random criticism and praise for answering questions demonstrate that “while people are not suckers for calumny, they are for flattery — even from a computer.” On receiving praise, we remain largely unconcerned about its source or basis, and “a smile or a warm tone…further amplifies the effect.” Bottom line: “sincere praise and flattery were equally likeable” to experimental subjects. Conclusion: “Praise and flattery benefit the praiser.”
“Praisers are liked and critics are hated, right or wrong.” Experiments involving newspaper reports quoting one person criticizing another revealed astonishing results: readers responded with negative impressions of the critic, the criticized person and the newspaper. This led the author to conclude that “When criticizing, neither accuracy, inaccuracy nor simply repeating someone else’s negative remarks gets the critic off the hook.”
Such findings have huge implications for politics, and further experiments were done to check into this. Professors Diana Mutz and Byron Reeves did an experiment replicating two politicians arguing on a talk show, and concluded that when the speakers were highly critical of one another, “people better remembered which candidate was on which side of each issue than when the candidates did not attack each other.” Logically, this should lead to more accurate issue-based voting, but whether it does so is hard to determine.
Though I feel provisionally cheered by many of the findings reported by Nass and Yen, I’m unsure how to square them with the recent and sudden rise of political incivility and the obvious appeal of internet hate speech and negative clickbait. This book was published in 2010 — before the recent sharp decline in political civility and the proliferation of internet vilification fests. Perhaps, like the small percentage of BC drivers who cause the vast majority of ICBC claims, the apparent darkness on the internet is the “work” of a small minority making a lot of negative noise.
Other findings show that criticism, justified or not, takes more mental energy to process and lingers longer in the memory. This is further complicated by “retroactive interference.” Following a negative event, people often cannot recall what happened just before. This is “a common problem for technical support professionals attempting to troubleshoot.” On the other hand, “praise and positive events do not require significant cognitive resources,” and thus cause no retroactive interference with memory.
These findings have implications for situations like giving performance appraisals. Since “after a negative event our memory is actually improved (proactive enhancement),” the best strategy is to “present the negative feedback first and then the positive evaluation.” This “will bring people to attention in time to listen to your praise.” Research also shows that to be more effective, criticism should be accompanied by suggestions for how to improve. In short, the bad news should come first, accompanied by suggestions, followed by the good news, ideally a longer list. Since praise “rarely spurs people to greater heights,” it is worthwhile for those trying to motivate others to “deliver praise that makes [comments] memorable.”
These are general principles, but as always in psychology, there are further complications. Mindsets — fixed and flexible. Self-evaluation (easier said than done). When things go wrong — blaming oneself and others. Personality — dominant versus submissive; cold versus friendly; clear versus ambiguous. Emotion: negative versus positive; excited versus calm. Expertise: real versus perceived. Computer tribes: PC and Mac, and the attitudes of their users.
There is much to learn from the experiments reported in this book. Culture. Trust. Stereotyping. Reciprocity. Speaking of reciprocity, this was perhaps my favourite question: Can a computer grant a favour and then say “You owe me one“? (BTW: just look at the cultural assumptions implicit in that.)
Life in society is a great experiment. We live through historic eras, each with its own popular mindsets. While we inhabit these mindsets, we breathe them in without awareness. A decade or five later, the same views are rejected, and seem outdated, misguided, even bizarre. The experiments reported in this book cast some light on human behaviour. This not only makes us aware of things that may previously have been influencing us while remaining below our radar, but suggests how we can use the knowledge to improve our lives.